Slashdot Mirror
E-Voting Glitch Alters Election Outcome
An anonymous reader writes "According to a local news source, 'A recently found computer glitch in the voting machines in Franklin County, Indiana has given a Democrat enough votes to bump a Republican from victory in a County Commissioner's race.' Any ideas on how we can check for similar problems in other close elections?"
It's nice to see that they're able to recount there, but it would be nice with an article about it that was longer than the /. summary.
hack diebold:
:= demvotes + 1
while (republican == winner) do
demvotes
repeat until (lawsuits stop OR
democrat(VictoryStatus) == true
Apply routine to all voting machines to achieve desired results.
Paper trail!
-- The morphemes of your disquisition are ascertainable, but they have eschewed an ambit of transpicuous exposition.
The voting software itself may be fine, but the instructions given it by the person in charge may be wrong. Don't forget to investigate the person, not just the machines.
I don't think these electronic voting machine problems should be characterized as trivial "glitches". They are complete failures of the software, since the whole purpose of these machines is to accurately count votes. Would losing a few hundred database records at your company be considered a glitch?
By referring to these problems as glitches, the media are downplaying the severity of the problem. Regardless of the candidates, if voting can not be reliable and verifiable people lose trust in the process and the outcomes will always be questioned. We either want democracy in the United States or we do not. But using technology that fails in its basic function should not be acceptable.
That description is poorly worded. It clearly states that the glitch caused the Democrat to win the election. Is this supposed to be the meaning? Or is it really supposed to say that the glitch caused the Republican to win, and getting rid of (bypassing) the glitch puts the Democrat in victory? In this case, the Republican wins with the glitch.
You have to love the ability to recount. No matter who wins, I want the election to be FAIR. You can only assure this with a paper trail for a re-count.
Some areas use only computerized systems, and how do you recount when you have a recording media failure?
HexaByte - he's a square and a half!
Remember that human volunteers have a high chance at screwing up also. Most of the volunteers in my area are over 60 years old (yes I live in Florida... LOL) and had huge glasses and were kinda crazy... like remember Will Ferrell as Harry Caray on SNL? yeah anyways..
of course there were a high percentage of the voters that were like that too...
Anyways, the best perfected machine (read most accurate) for counting votes should be the one we use. It should be the 99.9% accurate reflection what the votes were.
So what I say is, how can we tell these closed source systems work to 99.9% accuracy? Oh we can't.
So we're just supposed to close our eyes and trust the outcome we see on TV? Oh we are... hmm ok.
Makes me feel all tingly inside!
Get paid to code OSS
In any county where there is a close race, check the laws on recount and find enough people to insist upon a recount. Should be done countrywide at this point, given the problems we've seen.
Why the whole freakin' country can't just go to a proven system like Oregon's mail in ballots checked by scantron is beyond me. If it's good enough technology for SAT tests, it's damned well good enough technology for elections.
SJW: a person who perceives an injustice, and while correcting it, commits a greater injustice.
The glitch in the machines recorded straight Democratic Party votes for Libertarians.
That's not a bug... it's a feature.
Beauty is in the eye of the beerholder.
My friend sent me this, I only watched like 5 minutes because it seemed like propaganda but take it as its worth.
Votergate - a 30 minute video about the evils of electronic voting. The gist of it was bad computer, bad.
There are a number of voting machine-related challenges on the national level. Ralph Nader has successfully requested a recount in New Hampshire, and groups like BlackBoxVoting are working on fraud audits. Also, in Ohio, the Libertarian and Green Party candidates are reportedly joining together to demand a recount. There are local challenges going on as well. {Jonathan}
-------------------
Prof. Jonathan I. Ezor
Assistant Professor of Law and Technology
Director, Institute for Business, Law and Technology (IBLT)
Touro Law Center
300 Nassau Road, Huntington, NY 11743
Tel: 631-421-2244 x412 Fax: 516-977-3001
e-mail: jezor@tourolaw.edu
BizLawTech Blog: http://iblt.tourolaw.edu/blog
What is a "straight-democratic party vote"?
It take more faith to believe in evolution than it takes to believe in God
The purpose of an election is to collect and count the votes. Anything less than absolute accuracy is, or should be, completely unacceptable. Anything less that total transparency is, or should be, completely unacceptable. The process should produce enough documentatary evidence so that any disputes can be decided, without any doubt at all, in any court of competent jurisdiction.
Rigging an election should be a capital crime.
It isn't about red team vs. blue team, or sore losers, "desired results" or any of the other nonsense that is being thrown about to cloud the issue. I happen to be a republican, but I'm adamant about wanting this looked into. Why? Because honest matters more to me than "winning."
The way I was raised, if you cheated you didn't win, no matter what the score board says.
I have yet to hear a rational reason why anyone should oppose doing whatever it takes to make sure elections are fair, unless they are either cheaters or suspect that their side cheated and value victory more than integrity. What bothers me is that there are so many people in both parties that seem to fall into the later category.
-- MarkusQ
The text:
"A recently found computer glitch in the voting machines in Franklin County, Indiana has given a democrat enough votes to bump a republican from victory in a County Commissioner's race.
The glitch in the machines recorded straight Democratic Party votes for Libertarians.
The votes were re-counted last night, by hand.
The company who made the voting machine is also checking into programming of it's equipment in nine other Indiana counties. "
---------------
Doesn't this sound contradictory to everyone? The machine accidentally counted straight democratic ticket votes as libertarian while accidentally giving the democrat enough votes to beat the a republican?
I realize what it says is that after correcting the glitch the democrat gets enough votes to beat the republican who was previously determined to be the winner, but man that was horrible wording.
It seemed like propaganda? Normally when someone says that, they mean that someone was trying to convince them of something that wasn't true for nefarious purposes. What do you claim wasn't true? And more importantly, what sort of nefarious purposes to you suppose people have for wanting to make sure that elections are fair, or at least not quietly rigged?
And the point isn't that computers are bad, but that trusting a machine that was programmed by someone you have no reason to trust to do a process that you have no way to verify, when you are not allowed to see either the code or the data, is foolish. When the stakes can be valued in the multiple billions of dollars and some of the third parties have fraud convictions on their records, it's a little worse than just foolish.
-- MarkusQ
It's essential that the process be trustworthy.
Not having a trustworthy election is how revolutions get started.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
"Any ideas on how we can check for similar problems in other close elections?"
Unfortunately, there is no other solution than manual recounts. Not only in "close elections" because how do you differentiate a "far" (not "close") election from a large "glitch"? The only solution is to always do manual recounts--or just always count the ballots manually in the first place, skipping the "e-counting" step altogether.
The only way to make sure the votes are counted correctly, is to have a group of people representing all of the competing parties to witness and take part in the actual counting of physical ballots, look at each other while counting, compare the results, when they differ start from the beginning, and finally agree on one exact result. We cannot trust electronic counting the same way, because no one can witness and observe the counting process, no one can see the electrons being shuffled to eventually form a final outcome, just like we can see the paper ballots being shuffled and counted by people observed and verified by other people.
It doesn't even have anything to do with the source code being open or proprietary, the system being secure or vulnerable or the hardware being robust or faulty. It has nothing to do with the system being trustworthy or "trusted." The point is that being able to observe and verify the entire process we don't need to trust anyone or anything in the first place. And it means that the only way to have a solid democracy based on an e-voting system is to always do manual recounts, which obviously makes the whole e-voting idea quite counterproductive, to say the very least.
Sincerely,
Pan Tarhei Hosé, PhD.
"Homo sum et cogito ergo odi profanum vulgus et libido."
I don't care if the voter-verified audit trail is paper or microfilm that the voter reads through a microscope, as long as there's no way to hack the voter-verification process and no way to hack the voter-verified ballot afterwards.
For better or for worse, this means either the original ballot or the voter-verified audit trail MUST be stored in a human-readable form. For all practical purposes, this means in print. In theory, it could be in a computer-readable form that the person "plays back" on an independent-of-the-recording-machine playback device before signing off on it, but that opens up a whole new batch of "do you trust the machine" issues.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
I agree with all your points save this one. I'm not expecting 100% accuracy, but I am demanding 100% honesty.
-- MarkusQ
Even paper ballots have a "failure" rate, due to voters not marking the ballot as instructed - e.g. circling instead of marking an "x" - and anbiguity in election law on whether such marks are legal votes.
If, say, paper ballots have a 3% throwaway rate and a 0% error rate of good ballots, but a technological solution has a 0% throwaway rate but a 0.5% error rate due to voter error and bugs in the system, then it is better than the paper alternative. It's still not as good as it should be:
The only errors in voting should be voter errors, and if these errors are due to anything the state can prevent (e.g. confusing instructions), the state should prevent it.
No matter what you do, out of every million voters, some will simply mark the wrong choice and not realize it until it's too late. That's just tough.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
The glitch in the machines recorded straight Democratic Party votes for Libertarians.
This sounds like the core software was fine, only the configuration file for that election was erroneous. No amount of OSS on the platform level can catch the problem of misuse/errors at the election level. Even a paper receipt, scantron, punch card, etc. is no guarantee for forestalling this type of mistake. It's too easy for someone or something to misinterpret a mark on paper or in a computer file because of a miscommunication in the format, layout, or semantics of the ballot.
Two wrongs don't make a right, but three lefts do.
A simple search of Google News reveals it was a optical scanner, not a Diebold touchscreen system. Of course, if it had been a Diebold system, we wouldn't have this problem. No one would know the results were screwed, and no recounting would be possible.
l e? AID=/20041116/NEWS01/411160333/1008
4 21 -098.html
URL:http://www.pal-item.com/apps/pbcs.dll/artic
URL:http://www.indystar.com/articles/1/194039-4
Unless they are improperly calibrated or mis-programmed, most vote-counting machines are very very very reliable with correctly-marked ballots. They use technology similar to what banks have used for decades.
The problems usually come with ballots that are not clearly marked. If there's 2,354,365 votes for A, 2,354,301 votes for B, and 3,123 votes that got kicked out by the counter as too-hard-to-read, you can bet there's going to be a political cat-fight over hand-counting those 3,123 ballots - if the voter circled names instead of filling in the computer-readable fill-in bubble, does the vote count? What if he made a big circle that included parts of 3 candidates' names but 1 of the 3 was in the middle of the circle? At what point is the voter's intent "no longer clear"?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Software projects should be tested for correctness to the extent necessary to keep the customers happy.
For voting machines, this means they should be tested with correctly- and incorrectly-marked ballots in many combinations.
Furthermore, during the first years in the field, a sampling of elections should be partially (e.g a random 5% of actual ballots used) by hand and again by machine at the expense of the manufacturer, to validate that the equipment is still doing its job. Of course, the manufacturer is entitled to build this "post-sale quality control" cost into the cost of his equipment.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
They are complete failures of the software, ...
These are system failures. The entire workflow and resulting system design is plagued with deficiencies that many have reported. The software is only a tiny part of the problem. And, while e-voting greatly increases the number of potential failure points (many of which aren't software related), it's not just about e-voting. We have moved more rapidly to e-voting because of an equally bad paper-based design (punched cards with poor visual layout), but an election can also turn on something as seemingly trivial as washable thumb-print ink in Afghanistan. In every one of these cases, the state of the art at the time was much better than the poor systems that many people actually got. The major problem as I see it, at least in the US, is lack of pressure from vigilant voters on decision makers who should know better.
Mail in ballots are subject to fraud, intimidation, and even theft.
Can you imagine an control-freak head of household stealing all the ballots that come to his home then forging the signatures on each? You can't do that in a voting booth. Sure, family members could go to the police, but would they?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Not much information to make a story out of. Are they saying that a Rep should have won, but a Dem won instead because all Lib votes were counted for Dems?
.sigs are for post^Hers.
I used to live in Michgan.
You had a voter registration number.
When you voted, they gave you a ballot that had a number on it.
I think they recorded the ballot # with your name.
In practice, this information was kept far apart enough to ensure a secret ballot.
In principle, it was open to abuse. I didn't like it.
I don't know how they do it now.
This isn't about some small percentage of ballots being "spoiled" or some nebulous "voter error"; this is about the systematic miscounting of ballots, giving votes cast for one party to another. And it raises an interesting question:
If this is a bug in the software-as-certified, did it happen in every other machine of this make and model--which should have been identical? If not, why not? And how did software with such an eggregious error get certified in the first place? And if the bug wasn't in the software-as-certified, why and how did this machine come to be running uncertified software that systematically miscounts ballots?
These aren't the sort of things that can be explained away as "glitches"--they are examples of either fraud or gross incompetence on someone's part, and given the stakes I'd doubt gross incompetence.
We aren't asking for a perfect system, but we can quite reasonably demand an honest one.
-- MarkusQ
no one has cared in the past, no one cares now
I beg to differ. People have fought and died over this very issue. Perhaps you honestly don't care, or perhaps you just wish that others did not. But the fact of the matter is that the importance of honest elections may be the one issue that almost all Americans agree on.
-- MarkusQ
The old lever-system machines were notoriously inaccurate.
Any counting system will have tabulation errors, the question is, will they be 1 in a hundred or one in a million?
Furthermore, will they be due to design or manufacturing error (e.g. a gear that consistently slips in one direction on all units, shifting the outcome in the same way, or as in this case, a coding error) or wear-and-tear error (e.g. a gear that wears out on one machine, causing mis-counts or mis-recordings).
It's my understanding that the pull-lever machines had a paper trail of sorts, for use in recounts.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Many small elections are done with 100% accuracy.
If it's a local tax-increase election on paper ballots, it's not hard at all to see that 123 people voted for the increase and 124 voted against, with zero unclear ballots.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
In a close hand-counted election, where lots of eyes representing all candidates are on all the counters at all times, it's quite hard to pull off voter fraud.
Massive voter fraud using paper comes about because of a lack of checks and balances, not because of the technology used.
The same can be said for e-voting. One of the checks and balances missing from most e-voting is a voter-verified audit trail. Until that is fixed, it represents an opportunity for fraud.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Diebold OWNS the company that produce(d) these optical scanners.
We need non profit & organized voting standards. If corporate america can stand behind ISO standards why can't the federal government do the same?
If states require the rights to decide individually the votes (and laws) they cast for federal offices i'm not sure we can ever have a trustworthy system in the foreseeable.
I also believe we should streamline voting and make sure the right is protected and if people vote illegally it is punished for the crime it is. Partisan vote police shouldn't be allowed.
"Incidentally, I find it interesting that practically every EVM story seems to favor the Republicans"
This story does not favor the Republicans. The glitch itself did, however.
What's more interesting is that the old pull-lever machines had confirmed cases of intentional errors--election rigging. What's even more interesting, Ransom Shoup was convicted in 1979 of conspiracy and obstruction of justice one of these cases. He's the CEO of Advanced Voting Systems, one of the fine voting machine companies that provided equipment used in the most recent election.
-- MarkusQ
But the guy on TV said it was all OK. Those people complaining about the voting machines are just sore losers. At least, that's what I think he said. It was the guy that does the news right before the show with the girl who swears a lot.
-- Joe Average
According to USA today
So it must be the name of the county, not the technology, because the machines are from different manufacturers. Errm, yeah.Lars T.
To the guy who modded me down from perfect to terrible Karma - Apple haters still suck
But in this case, it came from software systematically giving one party an other party's votes. I don't think this is "voter error" unless the error is in blindly trusting our election officials.
-- MarkusQ
But, let's face it: it's harder to alter both the computer AND paper records identically than to just do one or the other.
Two scenarios, then:
1. Honest computer glitch gets discovered when paper ballots don't match up;
2. Dishonest computer manipulation gets discovered when paper ballots don't match up, although paper ballots aren't necessarily correct, either.
If you take the position that most (if not all) of these issues are honest glitches (as the emachine defenders often do) then you should be thrilled to have paper trails, as they'll uncover the glitches -- just like what happened in this circumstance. Really, it's delightful to see what can happen with a paper trail backup, isn't it?
On the other hand, if you know that the "glitches" are usually manipulation -- then you're probably going to avoid paper trails like the plague.
It's bad enough when I see it from AOL n00bs, but a news article?
Omnes arx vestrum sunt adiuncta nobis.
To put a finer point on it, what would you call an "error" in banking software that systematically deposited money into the wrong persons account? A glitch? Or what about a spyware program that consistently failed to report one particular company's spyware?
It isn't as if this software "failed" in the usual sense of the word--which implies that no benifit accrued to anyone. They didn't spit out error messages. They didn't burst into flames, or lock up. Instead, they superficially appeared to work perfectly but in fact were secretly highly biased.
-- MarkusQ
Comment removed based on user account deletion
From the Indianapolis Star website, Glitch causes Franklin Co. recount
Soylent Green is peoplicious!
period. This is exactly why and I can't justify the wasted time of standing in line with people who don't use deodorant to vote for a guy who I feel will do the least amount of damage to the country. Call me pessimistic, but hey, c'mon...
-- Game Developers: Stop porting badly-textured games from crappy console systems!
Support the Iraq genocide
Stop the pacifist plague. Ask your representative to bring home US criminal heroes.
The Guardian - Marines defend soldier's killing of Iraqi
UN News Centre - Iraq: UN human rights chief concerned over plight of civilians in Falluja
Amnesty International - Iraq: Urgent action needed to prevent war crimes
The more appropriate question is "Why is it that the glitches that favor the republican party are the only ones the media talks about".
There have been plenty of glitches that hurt the Republicans, most notably in Carteret County, N.C. where 4,500 votes were permanantly lost. Gone. Not recoverable. No recount. This county has historically voted 65% for Republican candidates, so this "glitch" cost Bush almost 3,000 votes.